The goal of this application is to further characterize immune responses that can cause the elimination of recombinant adeno-associated virus (rAAV) transduced cells. Most humans are naturally exposed to AAV-2 together with a helper virus and thus have immunological memory to AAV-2. Memory T cells can be triggered more readily than naive lymphocytes, which was not taken into account by pre-clinical animal experiments conducted thus far. Concerns about immunological consequences of rAAV-mediated gene transfer were substantiated by the outcome of a clinical trial in which human hemophilia B patients were infused into the liver with rAAV-2-F.IX vectors. Only one patient achieved therapeutic levels of F.IX, which were sustained for 4 weeks and then started to decline. At the same time the patient developed a transaminitis, which resolved after F.IX levels decreased to baseline levels. Overall, the patient's clinical course was compatible with immune-mediated destruction of rAAV-transduced hepatocytes. Additional data generated since substantiated our hypothesis that AAV-2-specific T cells induced by a natural infection can cause elimination of rAAV-transduced hepatocytes. To further define immune responses to AAV and rAAV-encoded transgenes under conditions thai mimic those of human hemophilia patients and to then devise informed strategies to circumvent such problems we are proposing 4 interlinked Projects supported by 2 Cores. Project by High will define T cell responses to AAV capsid proteins in humans and in non-human primates and assess their effect on hepatic rAAV-mediated gene transfer. Project by Ertl will elucidate the effect of pre-existing AAV-2-specific T cell-mediated immunity on hepatic rAAV vector-mediated gene transfer in mice. Project by Herzog will define regulatory immune responses that prevent induction of CD8+ T cell responses to a rAAV-encoded transgene product. Project by Terhorst will focus on the potential use of regulatory T cells to ablate unwanted immune responses to rAAV-mediated gene transfer. The projects will be supported by an Administrative Core, which will provide the needed oversight, and a Vector Core, which will provide the investigators with purified and quality controlled vectors. [unreadable]